Heavy Job Load on Argon

This week we've received several reports from community members that Argon is generally slow, in particular scheduling jobs and responding to file access. We've been monitoring the system and believe this is simply due to heavy usage with an unusually large number of active jobs.

Please note that jobs run in all.q are lower priority and are evicted to run jobs in the higher priority UI and investor queues. If you have a large number of jobs to run, and particularly if the jobs are relatively short-duration, please observe these best practices:
https://wiki.uiowa.edu/display/hpcdocs/Best+Practices+for+High+Throughput+Jobs

We may contact users about modifying problematic jobs, and we occasionally implement technical limitations to safeguard against particular problems. We apologize for the inconvenience and appreciate everyone's patience! Feedback and problem reports are always welcome, and we're happy to work with users on any specific issues which arise.